00:00welcome we have Eric here today so Eric
00:03is a YC partner he was previously the
00:06CEO and founder of pebble which went
00:09through the YC batch and all the way
00:11winter of 2011 a long time you're gonna
00:14ban Eric so now we have the other
00:17building on the other side but my Y C
00:19batch was one of the last ones in this
00:21small small building yeah you actually
00:24it was called Alberta when you applied
00:27yep and our product was called impulse
00:29impulse yes we had a whole bunch of
00:32complicated name how did you I mean
00:34pebble is a great name how did you get
00:36to a much better name yeah I think the
00:41original names were found using a
00:43thesaurus or some sort of like we just
00:45like scan through a book for words that
00:47were related to like alerting alert ah
00:51that was how we got to it yeah
00:52originally peple was much better name I
00:55actually thought of it while lying on a
00:57beach and it was like two years before
01:00we actually made pebble I just like
01:02wrote it down in Evernote on a list of
01:04good names nice you never know you have
01:08no idea when you're gonna use another
01:10name so any good ones just written down
01:13so today I want to spend a little time
01:15talking about hardware startups giving
01:18you ran one especially one that was one
01:21of the larger hardware starts in the
01:22past few years but then spend a bulk of
01:24the time talking about hard tech because
01:26you spend a lot of time working with our
01:28tech or here at YC and there are a lot
01:32of unknown still out there and we
01:34obviously want to fund more of those
01:36couldn't startups so starting with
01:39pebble so can you take us through like a
01:42quick story up you know when you came to
01:44I see how big were you what did you
01:47doing well I see like that's actually a
01:48common question of what hard for
01:50startups actually do what I see in three
01:52months so we applied to YC started the
01:54company in 2008 started working on it
01:57while I was still a student at Waterloo
01:59kept working that through 2010 we hadn't
02:03actually shipped or we were we were just
02:05starting to ship the product to first
02:07customers this is before couple back
02:10into in mid-2010 and
02:13I've been on Hacker News forever but I
02:15never really understood what Y
02:17Combinator was I just thought it was
02:19like the domain name for Agri news and
02:21one day someone said hey you should
02:23check out like YC and so just on a whim
02:26we applied the video we recorded if I
02:31look back at it was during Oktoberfest
02:33in Kitchener Waterloo where I was living
02:36and we were wearing like Oktoberfest
02:38Macedon video we applied for some crazy
02:42reason Paul Graham Jessica and the gang
02:44decided to accept us even though I think
02:47we were the third hardware company that
02:49we started YC in 2011 in the winter and
02:54the company the stage of the company was
02:56we had shipped maybe take two or three
02:59hundred of our very first watch the
03:01watch that that was called impulse that
03:03worked with blackberry smartphones and
03:05we had some early customers we were
03:07getting money but we hadn't we were hit
03:10we were earning some money hadn't raised
03:12much I think I borrowed fifteen thousand
03:15bucks from my parents got some
03:17government grants in Canada which was
03:18quite easy but we hadn't raised any
03:20money from investors YC was the first
03:23first investor so $15,000 let you
03:27manufacturers at the first one hundred
03:29units I we ended up making them had one
03:32by one actually so the first watches for
03:35this first watch they were manufactured
03:38we bought the circuit boards from China
03:40we had the metal the kind of metal
03:42casing of the watch made at a small
03:45metal shop a guy had a CNC a five axis
03:48CNC printer or a mill and he actually
03:51just like he read about us in the school
03:54newspaper or something like that and
03:56offered to like be our first shop for
03:59making the actual metal cases so the
04:01middle case would get made just to write
04:03like couple blocks away you get the
04:05circuit boards shipped in from China we
04:07also made the plastic watch straps in
04:09China and then we would do the final
04:10assembly one by one in the garage and I
04:15think that early process of making
04:16everything one by one was perfect for us
04:19because we screwed up everything you
04:21could possibly screw up the first ten
04:23watches we had weight
04:27a lot of problems but we had problem
04:29with RF we couldn't get the signal out
04:31of the watch because it's like a Faraday
04:33cage sitting on your wrist and so we
04:35said well maybe let's change the metal
04:37back for a plastic back and so we had a
04:40whole bunch of laser-cut acrylic you
04:42know she got a plastic cut to the same
04:44size as the back of the watch and we
04:46kind of jammed it in and then we shipped
04:49the first ten of them out by FedEx and
04:51on something like seven out of the ten
04:54the electronics just exploded at the
04:55back of the plastic and like people
04:58would show us pictures of like this
04:59watch the vigeous Bob with like the guts
05:01just everywhere and so we adjusted the
05:04thing we started putting like up like a
05:08piece of foam rubber inside the watch so
05:11when we shipped it the the backswing
05:12explode but it was a really good thing
05:15that we were making them one by one or
05:16else we could have potentially made like
05:18hundreds with that same problem right at
05:20the beginning and had lost you know
05:22so we I think we iterated really really
05:25quickly and we screwed up it was only a
05:27maximum of like a pretty small screw-up
05:31and how many units were you that you
05:34sold before you actually went through
05:36the entire manufacturing process we
05:38continued to manufacture hand by hand
05:41one by one by hand up until around 700
05:47so even when so one of the things that
05:49people ask about YC is like well what
05:51happens if I have like have to move my
05:53business to to Mountain View you know
05:55it's gonna interrupt my production
05:57process or whatever so what we did was I
06:00hired a couple high school students and
06:03I asked one of my friends who I went to
06:05school with who was doing his master's
06:06back at Waterloo just kind of like
06:08manage and run the the production thing
06:11and they would ship him down here and we
06:13would finish it in our house that we
06:15were all living at and so I think the
06:17trick there was like we didn't we didn't
06:20let the fact that we had to be here in
06:21Mountain View interrupt our actual like
06:23sales process and so after YC pebble
06:28obviously took off and lots of great
06:30things happen more of it I mean he went
06:31through lots of ups and downs it's a
06:34year after YC it wasn't like immediately
06:35after took a long time yeah so what was
06:39part maybe out of your old journey with
06:42pebble looking back I think it's
06:46different because you know seeing it
06:48forward it actually took us a long time
06:51to get from the initial idea to any sort
06:55of like real product market fit it was
06:57five years so we started the company in
06:592008 built the initial prototype late
07:022008 2009 since start shipping until
07:062010 and that was for our first watch
07:08and then we you know had a lesson had a
07:11lot of lessons learned from that first
07:12production run and then we redesigned
07:15and created what would become pebble in
07:172012 so that's five years and the
07:22hardest thing like I would say we work
07:25with a lot of founders and the hardest
07:27question that I think I ever get as
07:29someone who helps startups is like how
07:31do I know that what I'm working on is
07:34the right thing to work on how do I know
07:36that this little pivot that I could do
07:38is not the right one like maybe I should
07:39try that and there's no answer to this
07:41it is the hardest problem and so I you
07:44know thinking back I think it was
07:46probably very misguided in that I just
07:50had faith and I just never gave up even
07:54though now thinking back it's insane
07:56doing something for five years
07:57like what's the definition of insanity
07:59doing the same thing and hoping for
08:00different results like I was basically
08:02doing that for five years like we were
08:04just I guess we didn't know what we
08:06didn't know like we were very naive and
08:08we had a very small amount of money and
08:11we didn't know that you're not supposed
08:13to make consumer hardware products on a
08:16budget of $40,000 like most electronics
08:19that you use are made with budgets of
08:22millions of dollars I mean but now that
08:24you've done it or you did do it do you
08:27think it's possible now it's todya
08:29hotspur who know what I'm saying is I
08:31think like the hardest part that we had
08:32to go through was not letting other
08:35people get us down and kind of like
08:37maintaining that naive it say one of the
08:39things that I really love about like
08:40that phase at startup was people
08:44couldn't really convince us that what we
08:46were doing was wrong because no one had
08:48really done it before and if people said
08:53you should you know and you should look
08:55at raising more money or something like
08:57that me did really have the resources to
08:59do that like we were very we were very
09:00constrained and naive team and so he had
09:03to figure it out and I think some of the
09:05ways that I could screw it up was
09:07actually the ways that I did screw it up
09:09was right after YC we raised a bit of
09:13money about two hundred and fifty
09:15thousand dollars including this this
09:19kind of large check that we got from one
09:22investor and I took all that money and
09:24invested in inventory because I thought
09:26at the time the biggest constraint on us
09:28growing was we didn't have a reliable
09:31like we didn't have a large amount of
09:33imagery we were making them by hand and
09:35so we only had like 10 or 50 a week and
09:37I was under the mistaken impression that
09:40if we could start selling it in real
09:44time rather than saying we'll do a
09:46pre-order in celle - in a couple weeks
09:48that our sales would have taken off more
09:51because then we tried this we
09:53manufacture we instead of making them by
09:55hand you know in our garage we went to a
09:57small contract manufacturer in San Jose
09:59and manufactured 3,000 at once and spend
10:04all of our money on that and you know
10:06you know I'll see that the story is
10:08going it takes longer than you expect to
10:11bring up a new manufacturer so it took
10:13six months or eight months instead of
10:14three months it cost more than we
10:16expected so the BOM cost Road rose and
10:18then by the time we actually manufacture
10:20them that eight months rolled around our
10:23email list of potential customers had
10:26kind of dried up everyone had switched
10:27from blackberry to using iPhones and
10:30then we were stuck with a whole bunch of
10:32inventory so that yeah that was a really
10:34that was a really rough yeah difficult
10:36part yeah it's often not always but in
10:40many cases people get really creative
10:43when they don't have much money it may
10:46have to get really scrappy and then when
10:47they raise a ton of money they forget
10:49how to do that later
10:53we raised a bunch of money and then
10:55started hiring it and this is a big
10:57problem even mostly the successful
10:59companies run into and so I think the
11:02moral of the story for me at least how I
11:04think about you know
11:05that I work on is like that that
11:08frugality is vital to me not everyone
11:13has the same and not every industry is
11:15exactly the same if you crazy industries
11:16right now which are kind of by
11:19definition the opposite approval I'm
11:20thinking of like the scooter scooter
11:23sure where it's a matter of like raising
11:26as much money and it's physically
11:27possible but for the vast majority of
11:31people like staying frugal is what
11:32actually like yeah if you don't have
11:35Network like obvious network effects our
11:37first mover advantage then usually you
11:39should get on the other path so pebble
11:42is known as one of the most you dip them
11:46one of the most successful crowdfunding
11:48campaigns on Kickstarter and so a lot of
11:51people view crowdfunding is a way in
11:53lieu of fundraising to do that route so
11:57I think there are a lot of questions on
11:59what is the right time to do it
12:00especially like at what point do I need
12:04a you know a product that is real and it
12:07do I need to align that with the
12:09crowdfunding campaign and so forth so
12:11what's your advice on that I have a
12:14after having gone through a couple of
12:16these cycles here's my advice
12:20I think crowdfunding can be useful for
12:22two specific points I mean mainly useful
12:26for two specific points in a startups
12:28like you could use crowdfunding at the
12:30very early stage when you have an idea
12:33on the back of a napkin maybe a render
12:34or two to raise twenty five fifty
12:38thousand dollars that can help you get
12:39to a working prototype this is for
12:44people who are software engineers and
12:45are experimenting or want to work with
12:47hardware this is for hardware engineers
12:49who have zero money and they just want
12:51to like do they just want to prove that
12:52some people want their product so I want
12:55to be very specific like this is it this
12:57is an intention play it's not trying to
13:00be to build like a multi-million dollar
13:02campaign with like super fancy CGI
13:07graphics and overproduced videos like
13:09I'm talking about people honestly saying
13:13I just need a little bit of money to get
13:15this idea off the ground and I think
13:17there's an opportunity to use crowd
13:19right there it can be as simple as just
13:21emailing a bunch of your friends and
13:23saying hey do you want to just want one
13:25of these this is like the best test that
13:27you could that you can have for early
13:28product market fit is do your friends
13:31and family actually want to pay you
13:32money for what you're making or do le
13:35customer to actually want to prepay for
13:37what you're making it's not perfect
13:38because you know they may want to try it
13:40first or something like that but if you
13:42are making something that people really
13:44want they're basically throwing their
13:45money at you so that's it that's an
13:47opportunity for kind of this early
13:48crowdfunding I really encourage people
13:51to not try to do the marketing blow out
13:53campaign at the start for two reasons
13:56one is just looking back most of those
13:59campaigns even if they do succeed at
14:00raising money usually fail to produce
14:03and actually ship their product and I
14:06think the reason why is going through
14:07this cycle of having to make something
14:09and ship it to customers is really
14:12educational and you also learn what
14:16people like about your product or what
14:18they don't like as quickly it's like you
14:20reduce the amount of time between you
14:21having the idea and you're getting
14:23validation of do people actually want
14:25this idea so that's the first I think
14:27opportunity for crowdfunding I think
14:28there's actually another opportunity
14:30using it more as like a marketing and
14:32sales channel which I think more of what
14:34we did where we had already made most
14:36people don't know this but when we
14:38launched our Kickstarter campaign we had
14:40already made and sold around 1500 of our
14:43first watch and so we had gone through
14:45like all the pain and the cycle of
14:47building that and we had a really good
14:48idea of what pebble should look like
14:50because we had feedback from about 1,500
14:54individuals who were like super eager
14:57early adopters very open with their
14:58feedback and we tune and tweak the
15:01product so I think there's kind of like
15:03two separate crowdfunding opportunities
15:05and like a consumer products life cycle
15:07I think too often people blend them
15:10together or they don't know which path
15:12they're on and so they kind of screwed
15:14it up they they there's actually a
15:17couple other platforms for this early
15:19crowdfunding there's you know
15:20Kickstarter and IndieGoGo and that kind
15:22of thing but there's also tindy there's
15:24crowd supply there's like hackaday
15:26forums it's a lot of like really niche
15:28areas where you could you go to a
15:31discord server and say
15:32like you don't have to use the
15:34Kickstarter platforms for that early
15:37phase do you suggest having like a date
15:42in which you know I can promise you I
15:45can actually deliver this because if I'm
15:48using Kickstarter with the first with
15:52you know the first route maybe I don't
15:55know that I can deliver it within six
15:57months is there's like a certain time
15:58period in which I know I can deliver
15:59then I do that and if you're not like I
16:02would I would encourage like if you're
16:03not confident in your ability to ship it
16:05then don't use one of these like
16:08crowdfunding campaign platforms where
16:10it's very centric like shipping centric
16:13there's other options and again like I
16:19really do encourage going through that
16:20first cycle to find out things like how
16:22long does it take me to make things do I
16:24have all the skills on my team or do I
16:26actually have to find you know people
16:28potentially in the community to help
16:30contribute software or other other
16:33skills to the team again like too many
16:36people I see like they skip that step
16:38and they try to go directly to like the
16:39high-powered marketing campaign and it's
16:42just it's just such a sad time most of
16:45the time yeah and sometimes there's on
16:51Kickstarter kind of early adopters yeah
16:53but there's a group before now exactly I
16:55think for your given industry and for
16:57your given product category you should
16:59be able to find it like sometimes it's
17:01as simple as going on like a subreddit
17:02like we've had some great companies come
17:05through YC and they do their initial
17:07sales just by earning credibility in a
17:10Facebook group or they earn credibility
17:11in a reddit you know group or discord
17:13like there's a ton of customers there so
17:17when you talked about pep Laura Lee
17:19talked about hiring people and
17:20outsourcing things what's the right call
17:22in terms of outsourcing the various
17:26components so this is a really hard
17:28question for hard tech companies because
17:31more often than not the skill sets that
17:33you'll need to build to build your
17:36products are actually quite varied you
17:38may need a mechanical engineer an
17:40electrical engineer a computer engineer
17:45production people ops people and you
17:48don't necessarily need 100% of that
17:49person all the time you need like 15% or
17:5220% for three weeks and then you're back
17:56to not needing an electrical engineer so
17:59the best the best way to solve this is
18:01to have really flexible technical people
18:04on your team where you guys can just
18:09kind of incorporate everything and just
18:10do it yourself because those teams move
18:12so much faster than teams that require
18:15an external kind of process some of the
18:19most painful moments of my early career
18:21were not having full-time co-founders
18:26and having to work with contractors or
18:30consultants who honestly like you can't
18:34pay someone enough to care I mean
18:37firstly because you probably don't have
18:38that much money but also because people
18:41who are working on contract they're just
18:43set up to not have the same motivations
18:45than the same alignments that someone
18:47who's working 24/7 trying to get their
18:49fledgling startup off the ground has you
18:52run into time commitments I just had got
18:54an email the other day for my founder
18:55who's like yeah our intern went back to
18:58school and he was developing a large
19:00portion of our company and so yeah so
19:05like I actually opted for in the early
19:07days and pretty much all days trust and
19:10dependability and and and kind of
19:14smartness over very specific deep domain
19:20specific knowledge in who I'm looking to
19:22hire so last question on hardware
19:25specifically there's this common saying
19:28or adage what do you want to call it
19:30like hardware you can't just build
19:31Hardware you have to have a software
19:33component so that just scales you can
19:35have like a subscription service on top
19:37of it instead of just the one off type
19:39purchase how much do you believe in that
19:43or what do you think about people who
19:45say that all the time so I've been I've
19:48been working on I've been doing a bit of
19:49research into this category it's really
19:51got me interested right after pebble
19:54because that was some feedback that we
19:55often got could you launch a
19:57subscription service could you have some
19:59sort of payment subscription and you've
20:06heard of product market fit right you
20:08know the idea that customers really you
20:11know in a particular market are going
20:12after your product I actually think
20:14there's something called like business
20:16model company fit where you could have
20:20the best product in the world you could
20:21actually have product market fit but if
20:23you have the wrong business model you're
20:25not gonna be able to fund and you know
20:27grow your company over the long term so
20:30it's kind of like an advanced thing but
20:31you don't need to worry about this if
20:32you'll have product market fit because
20:34you know how product market fit you know
20:36you don't have a company but we ran into
20:39a problem pretty soon after we hit
20:41product market fit of misunderstanding
20:44what was the what was the business model
20:46that was going to drive this company and
20:49I think there's actually like seven or
20:51eight or nine different business models
20:53that are useful that that are kind of
20:55like lanes that hardware companies or
20:57heart tech companies can can use I've
20:59been working on the kind of just kind of
21:01researching and putting companies into
21:03different categories and you have the
21:04subscription companies like the nests
21:06and the or the dropcam's of the world
21:08that are selling you know ring healthy
21:11three to ten dollar subscriptions per
21:12month but then you see other companies
21:14that try to do something similar and it
21:16just does not work and I think that's
21:18because so in our case we didn't
21:21struggle I think with subscriptions we
21:23experimented with it it wasn't really
21:25going to work for us but the way that we
21:26screwed it up was we didn't understand
21:29that we were in hits driven market hits
21:32driven being you come up with a consumer
21:34product hopefully it's a hit it hits
21:36that product market fit and people start
21:38pulling it out of your hands similar to
21:42a game company we didn't realize that we
21:44had to be super frugal super kind of low
21:51budget as we developed the new products
21:53and kind of tried to engineer another
21:56hit our first product was certainly a
21:57hit I don't think we ever reached escape
22:00velocity to a hit with any of our later
22:02products and I think if we had built a
22:04more frugal organization if we were able
22:07to build products cheaper and then
22:09launch them and test to see if they were
22:13they were successful scale it up etc I
22:15think we might have had a slightly
22:16different outcome and it kind of goes
22:18back to like there are these different
22:20models for hardware companies and you
22:22need to like you need to you need to
22:24make sure that you're on the right but
22:25you have to question yourself and you're
22:26like should we really be on a
22:28subscription model is this really what
22:30customers are gonna want what they want
22:31to spend five dollars ten dollars a
22:34month and you can't force it and if you
22:36force it like I think we tried to do you
22:39you might add extra problems like on top
22:41of just building great products for your
22:45so pivoting a little bit too hard tech
22:46so we've funded lots of cool hard tech
22:49companies lately quantum computing
22:51self-driving cars pretty printing entire
22:54homes so for agriculture technology yeah
22:57nuclear fission nuclear fusion aerospace
23:00space space yes Rockets satellites and
23:03rockets so I mean we can calibrate the
23:06conversation a little bit how do you
23:07define hard tech deep tech what does
23:10that really involve for me I think what
23:14what I'm excited about you know in
23:17harder technology spaces is that is the
23:21in the in the software world the
23:23iteration cycles are quite short
23:26you can push code to your to your repo
23:29and have it go live within seconds and
23:32so you can rapidly iterate on your
23:34product and get feedback from your
23:35customers I think hard tech companies
23:37can't they don't have this same rapid
23:39iteration cycle they either have
23:42hardware in the mix which kind of is not
23:45at that same speed yet though it is it
23:46is dramatically decreasing the amount of
23:49time that each iteration cycle takes or
23:51you might have some sort of fundamental
23:53technology that you're working on that
23:55needs to be developed in order to unlock
23:58a new business model or new use case or
24:01you might be regulatory regulatory
24:04constrained either in the medical device
24:07space drone space transportation space
24:11software on car space these are all like
24:13there's external processes that you
24:15can't control that affect your iteration
24:18cycle time so say this crazy idea
24:21something to do with rockets or
24:23something insane it's to most people
24:26normal to me of course rocket scientist
24:29and I have no money what away what do we
24:32even start what do I do it doesn't cost
24:36that much money to get started these
24:37days you can do something like startup
24:40school where they give you free money to
24:42invest in like if you're working on AR
24:45machine learning you can use that to get
24:46compute credits well maybe a better
24:51question is what is an MVP for most of
24:54these yeah Carter Tech so for medical
24:57I think the MVP is successfully
25:00peer-reviewed science that supports what
25:03you're working on that that there's not
25:05that there's some sort of opportunity
25:06here and a kind of understanding of what
25:10your path to clearance is those are in
25:14my opinion I think the best companies
25:15because they're the ones who have shown
25:16that they understand which business
25:18model they're in and they know what's
25:19going to take what's gonna be taken of
25:21them to get to kind of sales but they
25:25don't necessarily have to be like too
25:27far along that path I think at that
25:29stage investors start to get excited
25:31when they see the mating of science
25:33Senna and a use case we've had a bunch
25:37of companies that are working on rockets
25:39and satellites and propulsion and that
25:43for those cases it's usually some sort
25:47of subset scale representation of what
25:50you're working on and customers so in
25:53all cases even if you're working on
25:55regulatory constrained products having
25:57customers either contributing money to
26:01you through early contracts or or
26:05further along on the kind of pilot track
26:08the better I think oftentimes engineers
26:12we love to build things we don't as much
26:17love to sell them and so what I look for
26:19in early-stage companies is are the
26:22founders technically competent and do
26:24they understand that they're gonna have
26:25to kind of perverse themselves become
26:27salespeople or not what do you find by
26:31demo day when they're in front of lots
26:33of investors what is it that you
26:34encourage them to accomplish by then it
26:37sounds like there's two passers the
26:39if you can do sales there's no
26:41regulatory issues to doing sales you
26:43should have some paid contracts ll eyes
26:45maybe and then also a technical
26:48validation like proof that you can
26:50actually built this yeah and I think too
26:52often people convolve these and they say
26:54I don't want to do sales until I have
26:56something that I can show to them that's
27:02most customers if you explain that they
27:05have a problem and you have a way to
27:06solve it they'll get excited and if
27:08they're not excited it's probably a sign
27:11that they don't really have a problem
27:12that they want to solve and so
27:14oftentimes we get companies that come in
27:17at the beginning YC and they say well we
27:19want to spend the first two months you
27:21know working on improving the prototype
27:22and getting it to the point where we're
27:23proud of it and then we'll do sales in
27:26the last month and you know we kind of
27:28flip that around like do the sales hit
27:31the pavement like talk to customers and
27:33you'll actually build a better first
27:35product for them than if you just built
27:37it first and showed it it also gives you
27:42confidence that you're going down the
27:44right track someone said it's hard like
27:46we had a company in the last batch that
27:48had been working for two years on
27:51building a prototype they were
27:53commercializing some technology that
27:54they worked on in a university and it
27:58had a hardware component and I remember
28:00seeing I think was their third
28:01application second or third application
28:03to YC and they had kind of similar
28:06videos of them doing the demo with their
28:08technology and we you know we liked
28:11their commitment like they had been
28:13coming back and the technology was
28:14getting better and they were just
28:16starting to talk to customers and we
28:17said let's give him a shot and so we
28:18brought him in for an interview accepted
28:20them in and then basically the first
28:22week we took him aside and just said you
28:24need to start selling you need to get
28:27people like excited and interested and
28:29it was almost like a 180 for them
28:32because they thought that we were going
28:33to say you should work on technology and
28:34get to the point where you have like
28:36working units widgets that you can sell
28:37and we said no like start going and
28:41talking to customers and they it took a
28:43little while for them to get up to speed
28:45they you know it's a it's like
28:47exercising your new muscle especially
28:50for engineers and within
28:53span of two months they went through
28:54several rollercoasters of like having
28:56zero leads to having a whole bunch of
28:58leads to having them all collapse and no
29:00one's signing and then like six or seven
29:03weeks into YC they they they sense no
29:07and they were like we got our first we
29:08got our first contract and then they got
29:10their second contract they got their
29:11third and then like my demo day they had
29:13already had already like line on four or
29:16five million dollars in sales and that
29:19all happened in the span of three months
29:20because they've kind of you know stopped
29:22working necessarily on just building
29:24their technology they were still working
29:25like they had a three or four person
29:26team so they hey they were able to split
29:28it up but the focus that number one goal
29:29became how do we get sales so say I mean
29:35you do a lot of trips where you go to
29:37schools specifically you talk to
29:38scientists and engineers many and peach
29:40tea programs because what is what is
29:45your advice to them to make the jump
29:47which they consider a lot of them are
29:49thinking about commercializing it but
29:51maybe they think I don't know what they
29:53think actually but but what is your I
29:56guess went like one-two punch on that
30:00having started my company in a kind of
30:05university town versus somewhere here
30:10like Silicon Valley I felt very kind of
30:12protected there were a lot of government
30:14grants I was able to raise non dilute of
30:17funding and Plus that is we were able to
30:23use that funding to work on the project
30:24but the downside was there was no
30:26pressure to have some sort of product
30:29that customers would pay money for it
30:31wasn't really in the DNA of the startup
30:33kind of ecosystem that I was in most
30:36people were just entranced with working
30:38on their startup and not necessarily
30:39getting to that next stage of building
30:41kind of a profitable longer-term Corp
30:43and I think that that sometimes affects
30:45people who are in universities even more
30:47because the the the default is raising
30:51money through grants and working on
30:52longer-term projects whereas the special
30:55case is like spinning out a company and
30:58actually like trying to get customers to
31:01buy it and I think what I would
31:04encourage is like try getting out of
31:07just as an experience like or experiment
31:09don't don't say that you have to like
31:11dive in headfirst with no parachute but
31:13if you're working on some sort of
31:15technology that you think could have
31:16some sort of positive business impact
31:19like get out of the lab go to some
31:22conferences talk to people not not
31:24educational conferences like corporate
31:25you know sales conferences start talking
31:28to customers and see what people think
31:31and like that's that's the start and
31:34hopefully I mean if you're if you're
31:36lucky and things are in the right spot
31:37again customers will be like where's
31:40this been all my life like this is this
31:41is amazing this is exactly what we've
31:43been dreaming of I can't believe it
31:44finally came out of the lab and that
31:46should be the momentum that just like
31:47pulls you pulls you along but you have
31:49to get out to please try it so so you
31:55get to the point where you feel
31:57confident you have like some amount of
32:01interest in your product but to actually
32:06build it you can eat lots and lots and
32:08lots of money what do you need to show
32:11to investors so the best the best
32:16companies in my opinion are the ones
32:18that don't require a ton of money to
32:20work on or at least to get to some sort
32:22of MV MVP stage kind of like with all
32:26advice that once he gives it's usually
32:28applicable to 90% of the companies and
32:30there's always 5% on either side that
32:32are in a special case but even some of
32:36the best hard tech companies I've seen
32:38they don't have like a binary outcome
32:40where they need someone to insert a
32:43hundred million dollars and that's the
32:44only thing that can develop the product
32:49like take I mean I don't know anything
32:52about these companies in other than kind
32:54of press reports but take kind of magic
32:56leap versus oculus right magic leap
33:00raised lots of money and spent 11 years
33:04working on a product versus oculus which
33:06raised two million dollars on
33:08Kickstarter shipped a you know early
33:12development development kit that was not
33:16amazing but they iterated and they got
33:18through and now the oculus
33:21the new one that's like 400 bucks
33:23doesn't new computer like it's it's
33:26exactly yea the quest like that's
33:28exactly what people are dreaming of now
33:30and they're gonna have it within six
33:33months and they had to go through like
33:35six seven years of like painful
33:37iteration but they did it and yeah so
33:44last question then we'll take audience
33:47what's an idea just so curious what's an
33:49idea that sounded so crazy to you when
33:52you heard about it but then when you met
33:53the founder and they stepped through
33:56everything or like oh that's not so
33:57crazy this is totally possible
33:59relativity space so that was company
34:03that went through YC for three years ago
34:05not very long they wanted to build 3d
34:09printed rockets their thesis was it's
34:13not actually rocket assembly that cost a
34:16lot of money it's the testing phase of
34:17integration or if you have hundreds of
34:20different systems that have to work
34:21perfectly with each other when you
34:22assemble and build the final rocket you
34:24have to test everything select a
34:27different path than like SpaceX which is
34:28going for the reasonably T thing so
34:30their their thesis was if we can 3d
34:32print the entire rocket from the tip of
34:35the exhaust to the motor to everything
34:38in like one piece the integration time
34:41is zero because you just print it so
34:44they're working on this and it was just
34:46an idea I think they did a hot fire
34:47during YC I think they were working on
34:50it I don't know exactly how long they've
34:51been working on it for but yeah that's
34:54pretty crazy idea and to come through YC
34:57especially to pitch to investors that
35:00probably were not you know all space
35:03investors must have taken some guts but
35:07they did it they showed off their first
35:09three printed rocket they raised early
35:13funding off of that they started work on
35:16their larger products and raised a whole
35:19bunch more money I think they just
35:20raised their P or C last year they also
35:25they worked at previously at SpaceX
35:28origin but they didn't have many years
35:32of experience so it was
35:34actually craziness ends also they were
35:36extremely young and for this kind of
35:39thing - yeah thing in this and people
35:42people believed in their in their vision
35:44they demonstrated enough I think of what
35:45they were working on and I think it was
35:48actually interesting so I think they I
35:50don't want to say too much in case I'm
35:53wrong but I think they were working on
35:54the printer first and they were looking
35:56for an application for this massive
35:58metal 3d printed they'd built and they
36:00found space as the I mean it was their
36:02background and their passion so I think
36:04that's the one that they took it in the
36:06direction that they took it in I see
36:08this a lot with people who are working
36:09on really cool kind of fundamental
36:14I think PG actually had a pretty good
36:15tweet about this couple weeks ago he
36:16said like if you're in a situation where
36:18you're building something that's like
36:19mind-blowing ly cool and revolutionary
36:21but you're finding that customers don't
36:23want to buy it pivot your companies so
36:26you're actually doing what your
36:27customers would have done and just use
36:29your pioneering technology to do what
36:31they do faster better or cheaper and
36:33sell to their customers and that's kind
36:35of I think what relativity did is they
36:37were building this new 3d printer and
36:38they realized that we could just build
36:39rockets with this technology and we will
36:42be better than everyone else who's
36:43building rockets in that space cool so
36:46questions first so the question was how
37:13do you get your first customers how do
37:15you get one through 50 I talked about
37:20this a little bit earlier some of the
37:22best ways is to kind of infiltrate the
37:24community of people who want who you
37:27think will watch your product there are
37:30an unlimited place unlimited set of
37:33places on the internet where people talk
37:34about products and they talk about
37:36services they can range from you know
37:40where the whatsapp guy posted his app
37:43first in back in 2009 now he posted it
37:48which is like a whole bunch of like
37:50geeky business people who fly a lot and
37:52that's where you posted it and like
37:54think about that as your first as the
37:56first kind of user group for whatsapp
38:01it's travelers who want to talk to
38:02people all around the world who are
38:04tired of paying for overpriced like SMS
38:06bills and it's 2009 so people are just
38:09starting to get like smartphones and
38:11that kind of thing perfect like you
38:13couldn't have planned it better if you
38:15tried and so the best companies are the
38:17ones that just kind of get out there for
38:20us we got a lot of press in the early
38:23days by talking to bloggers so I like
38:26formed personal relationships with all
38:28of the BlackBerry bloggers at the time
38:30if anyone remembers CrackBerry
38:32CrackBerry Kevin yeah and so I would
38:34like fly out to conferences on like
38:36their cheap red eyes and just like
38:38instead of buying tickets to the
38:39conference I think we were just kind of
38:41like sneak around and get into the print
38:43like once at CES back in 2009 we just
38:46just walked into and gadgets trailer
38:49like Engadget had this big trailer where
38:51all their reporters were working and we
38:53didn't know anyone so we just walked in
38:55and then just started like pitching to
38:57people and ended up getting us through
38:59this like a picture of my wrist with a
39:01broken watch it didn't even work at the
39:03time but it's like so-and-so is working
39:05on a SmartWatch and that drove a ton of
39:07traffic in the early days was just kind
39:09of people's self identifying as this was
39:11something that that I wanted yeah takes
39:14a lot of hustle yeah things are much
39:33easier so the question is like we're a
39:35bunch of software engineers we have an
39:36idea that had includes software and
39:38hardware how do we get started on the
39:39hardware well the good news is that it's
39:42easier than ever I wrote I wrote a blog
39:46post that was on TechCrunch a couple
39:48months ago about how software engineers
39:51can kind of get started on this and the
39:53recipe is pretty simple find an
39:56off-the-shelf product that kind of
39:59what you want to do that's kind of like
40:01halfway there that you can actually just
40:04buy off Amazon or Alibaba or something
40:06because the amount of like the amount of
40:09time and money and tears that you'll
40:11save just by buying a product that's
40:13halfway to what you want to build will
40:17will do a couple things one is it will
40:19get you within two to three days or
40:21however long shipping takes an actual
40:23physical representation of what you
40:24think you want to build and you will
40:27begin to use it hopefully to solve the
40:29problem that you want to solve and you
40:31will learn like maybe I'm completely
40:32wrong and this often happens like the
40:35best intention product when you actually
40:37build it might not actually solve the
40:39problem that you want to solve so
40:40getting halfway there for like two
40:42hundred bucks or three hundred bucks
40:44plus shipping it's just so much better
40:46than weeks and months of prototyping
40:48just to get to the same kind of level
40:50the next step after this is to take the
40:5250% their product or 75% their product
40:55and hack it to do what you think is
40:58better either adding a little bit of
41:00technology onto it adding your software
41:02stack modifying it can be as simple as
41:05like repurposing the same technology and
41:08like it's changing that changing the
41:09setup a good example of this would be
41:12kind of a scooter companies they took
41:14Xiaomi scooters you know nine bought
41:17scooters and they just glued on cell
41:20modems that's it that that's how scooter
41:24sharing starts like cell modems and GPS
41:25plus scooter he didn't build this
41:27scooter themselves they bought it off
41:29Alibaba and they bought the modems from
41:32some company that was making modems so
41:34that's kind of the prototype I think for
41:36the next generation of consumer hardware
41:38as well as enterprise hardware we had a
41:40company two batches ago that builds
41:43cameras for neighborhoods neighborhood
41:45security and they used a Raspberry Pi in
41:47a PI cam and a 3d printed case and they
41:51sold those for like hundreds of
41:52thousands of dollars the hardware cost
41:55you know very small amount of money and
41:56they're just solving a real problem for
42:00end customers and people are paying them
42:02in exchange for that so yeah it's it's
42:05even easier than ever you can find
42:06people in Alibaba you can meet factories
42:12don't try to build it all yourself don't
42:14hire like an industrial design firm to
42:16build you some new design don't hire
42:18like an engineering firm that promises
42:20to do a custom design for you because
42:22that'll cost hundreds of thousands of
42:24dollars and it probably will be the
42:25wrong product versus spending 300 bucks
42:28find something meeting the factory that
42:30makes that current product and asking
42:32them to kind of make the changes that
42:34you want to make it depends so the
42:47question is should you move should you
42:49move closer to changing or factories
42:52when you are pre product market fit if
42:57your free product market fit and you
43:00have a product that you're iterating on
43:02like figure out a way to reduce the
43:04iteration cycles so that you get more
43:06chances to dazzle customers and to find
43:08you know that product market fit so you
43:11could solve that two ways if you think
43:12that moving to the factories would
43:14be--if will enable you to do faster
43:16iteration and you already have kind of
43:17sales channels set up that's cool if not
43:20and you don't really have that many
43:21customers then move closer to your
43:23customers so that you know the iteration
43:25can be done in software maybe not
43:27hardware and you're just like spending
43:29every minute in front of your customers
43:30so we have a couple companies that did
43:32that in the last batch and they were
43:34using drones to solve very specific
43:36business process problems and they
43:39basically just moved to where those
43:40businesses are took up like a motel a
43:43crappy motel room and turn that motel
43:45room into a like a drone factory where
43:47they were making and modifying their
43:48drone and that in that case you know was
43:52probably recall for them protection IP
44:03protection trademark protection is that
44:06kind of what you're interested so yeah
44:09patents that kind of thing so this
44:11question I think kind of touches on how
44:13do you build a moat for your company how
44:15do you make sure that once you figure
44:17out a beautiful product figure out who
44:19to sell it to that some other dude
44:21doesn't come along and just steal your
44:22idea so there's a couple ways that you
44:26the best ways in my opinion are
44:28companies that have some sort of innate
44:30moat where it's either Network effect
44:33where the more people that use your
44:34product the better it is for everyone or
44:36some sort of data a lot data lock-in
44:39where you use the product you generate
44:41data the davida happens to be within one
44:43ecosystem someone who builds a competing
44:45piece of hardware can't get that data
44:47easily those are the two best what do
44:51you think about that I think those are
44:55good I also think patent previously used
44:58to be very hard to get but now there's
45:00like we have a company called cognition
45:02IP I would just use them and you pay
45:05them some amount of money they use
45:07software to help you file so the the
45:09problem with patents is that they're not
45:12like if you're thinking about this
45:13problem specifically how do we block
45:15other people from copying our idea
45:16they're actually really difficult tools
45:19to use versus those other two things
45:21that I described that are actually like
45:22you control all the levers if you can
45:24build a product that has built-in
45:25network effect or built-in data's you
45:28know things you control that you don't
45:31have to rely on lawyers you don't have
45:32to rely on the government you'll have to
45:34rely on the courts to enforce your
45:36domain dominance but if you were if you
45:39rely on patents or that kind of thing
45:42you're basically saying I I I'm putting
45:47this money down as an investment but I'm
45:49gonna have to invest a lot more money
45:51later in order to use unlock that first
45:53investment so it's very rare that
45:55startups build a successful moat using
45:57patents this happens much more in like
45:59biotech so yes yeah I agree with that I
46:03think the I think building really great
46:06product is really just the only way to
46:08really build the moat I mean again for
46:11like 5% of companies there might be some
46:14sort of patent strategy be a 95 percent
46:15it's it's great products yeah I think if
46:18it's also easy to build people just copy
46:21it even if you get that and so it
46:23doesn't really no that's the thing you
46:24need money to enforce the patents you
46:25need time to enforce it and if your
46:28companies dead it's hard to enforce
46:50marginal markets are great so the
46:52question was what happens if I'm
46:54building a hard tech problem longer-term
46:56trying to solve a longer-term problem
46:58that first problem that I'm solving what
47:01if it looks too small honestly the best
47:04investors will want to see a small
47:06market because a small market at the
47:08beginning means you can identify who's
47:11in that market very cheaply versus
47:14having to boil the ocean saying well
47:15there's like two percent of all people
47:17in the world who are interested versus
47:19like eighty percent of this one group so
47:21it's cheaper and easier for you to find
47:22those customers the second thing is that
47:25small group of people will tell you
47:28quicker whether you have a good product
47:30or not because they know what the they
47:32know that they have a problem and you're
47:34telling them hey buy this thing solve
47:36your problem they can use it in big yeah
47:38that solves it it's good yeah problem
47:40solved they'll give you feedback whereas
47:43if you're trying to again target like a
47:45really large group with a vague product
47:46idea it's gonna be harder to know
47:48whether that person just was interested
47:50that week maybe didn't have enough money
47:51you know there's there's more reasons
47:54than you had a bad product if it's a
47:56really vague in large large market yeah
47:58just add on to that I think starting
48:00small is good but also having a road map
48:03to how it builds into something big as
48:06good as well for investors specifically
48:07and for yourself as well to know like
48:30feminine so in my company by chance we
48:34just like to it's an electronic
48:37treatment for reducing sweating and we
48:40we are so happy that people are so this
48:50cuz this questions related to like do I
48:52do I pick a recurring business model
48:54subscription business model hardware is
48:55a service or do I just sell the product
48:58so this kind of goes back to what I was
48:59talking about before which is making
49:01sure that you have the right you you're
49:03pretty sure that you're in the right
49:05business model track for your company so
49:09there are a lot of these different
49:11tracks they're not just hardware's a
49:12service not just consumer there's hybrid
49:14models there's insurance models there's
49:17you know advertising driven models
49:19there's a ton of different ways that you
49:20could look at it so it's not just one or
49:22here's what I would say on the
49:25enterprise side you need to figure out
49:28how much money you will save your
49:31customer or how much more money they
49:33will make and price your product
49:35accordingly like your product price to
49:37them will be less than either how much
49:39they're saving or how much more they're
49:41making and you should try to price it as
49:44high as possible to still let them have
49:46some savings but you should you know
49:49rightfully so earn some of the profit
49:51from that innovation that you have
49:53sometimes that may be a hardware as a
49:56service model it might be upfront and it
49:59really depends on kind of your customer
50:01at that point so it's really hard for
50:02anyone to try to give advice on that in
50:04that front and it's much more around
50:06like talking to customers experimenting
50:08you could do price experiments know most
50:11companies are so small that every time
50:14like a new person hears sorry every time
50:16they do PR new people who've never heard
50:18about it will hear about it for the
50:19first time and they won't know what the
50:21last price was so you can afford to like
50:23do that kind of experimentation
50:24especially at the beginning of your
50:44yeah how do you do hiring especially
50:47when you're scaling hiring is a mode
50:52that the companies in and it's a very
50:55distracting mode so I think like the
51:01mission of an early stage startup is to
51:02get the product market fit and it's
51:04definitely not to like hire a team no
51:07one gets points for hiring a team or you
51:09get a very small number of points for
51:10hiring an amazing team you get a lot
51:11more points for finding a product that
51:15has proper market fit so hiring in
51:18general is really like a touchy subject
51:21these there's no one size
51:22one-size-fits-all there some general
51:25advice that I have is what I mentioned
51:28before which is higher for trust and
51:30higher for reliability over hiring for
51:34specific domain knowledge this has
51:36really benefited me because right after
51:40we kind of hit on Kickstarter we were
51:43only two people at the company two
51:44full-time people and two interns and so
51:47we had like nobody and all of a sudden
51:49we needed to hire people and we could
51:53have spent you know three to six months
51:55hiring a 30 person team and won all
51:59these different you know people that we
52:01would need hire expert experts that have
52:03done it before we could've spent time
52:04doing that we would have delayed our
52:06initial production by probably three to
52:08six months however long it would have
52:09taken to hire that team so what I did
52:11instead was I just phoned up like some
52:14of my best friends from University who
52:16had been tangentially involved in the
52:18project and said you need to quit your
52:20job and just fly here and I think all of
52:24them we just hired seven people within
52:27three or four days like that and that
52:29worked because we had that momentum and
52:31we had found product market fit and so
52:32they were like that's cool I want to be
52:34part of this rocket ship I want to take
52:35off but we didn't have to do like three
52:38to six months worth of hiring we did
52:40three to six days with the hiring
52:42and I knew them I trusted them they were
52:44instantly part of a team because it was
52:46just like we were working on another
52:48school project again it seems like the
52:50best source pool is either people you
52:53met in college or from somewhere else
52:55you've had yeah yeah like Pete like too
52:57often people overlook like some of they
53:00worked at a job with two three years ago
53:03one of the coolest psych hacks that you
53:05can do is just look through them
53:07manually slowly look through your
53:09Facebook or LinkedIn and just look at
53:11every person's name and think is this
53:13person cool and technical and just go
53:16through that list and you'll probably
53:18find a co-founder you'll probably find a
53:20first employee like some long-lost
53:23friend from high school that's actually
53:24like a super cool developer you could
53:26hire her and just bring her onto the
53:28team like there's the second step people
53:30miss then is actually asking that person
53:32and giving a real offer there's always a
53:34people do this ambiguous dance of like I
53:36don't want you to say no to me and
53:38that's so I'm not gonna really ask you
53:40and so there's there's a there's a good
53:42book called the hard thing about hard
53:43things by Ben Horowitz and he talks a
53:46lot about like working with friends and
53:48the difficulties and benefits of working
53:51with friends so I read that book used to
53:53end up hiring or working with a bunch of
53:56your friends all right
54:13is there a lick question was for some
54:18projects problems they can be solved
54:19exclusively by software how do you know
54:21when to involve hardware in it and when
54:24in doubt take all the hardware out does
54:26it solve the problem that's the best
54:28place to start that's why I was kind of
54:29saying like how could you just buy
54:30something off the shelf and use it
54:32because you may need some hardware you
54:34maybe be camera you may need a
54:35microphone you may need a compute
54:37platform just buy a Raspberry Pi and
54:39plunk USB devices into it like don't you
54:41don't need to build anything these days
54:42some people think like but it's not
54:45going to be the right form factor or
54:46it's not going to you know do the exact
54:48right thing as what I think our
54:50customers need just try just reject
54:53build something hack something together
54:54and see if it at least
54:56almost solves the problem because like I
54:58was saying more often than not that
55:00initial test will prove that either you
55:04have a really good problem that really
55:06needs to be solved or nobody really has
55:08that problem and you don't need to now
55:10spend three months working on custom
55:12hardware to mr. because the whole user
55:25experience make difference so actually
55:29we are creating the whole graphic it's
55:32not least 182 people but it's really the
55:36product is great but middle so it's
55:41really bad so what is the best way to
55:44pass way so the question was we really
55:53complex product an AI psychotherapist or
55:57therapist that can that can help people
56:00how do we demo that or how do we convey
56:04that product to people I would hazard a
56:08guess that you know it's kind of a
56:10similar question to last one it's like
56:11are there ways that we can do this in
56:13software rather than hardware so if I
56:16were working on a project like that I
56:17might start by actually creating the
56:19assistant as an app first or maybe like
56:22a 3d you know unity character that
56:25approximates it that gets similar close
56:28to what you want to do or use an
56:31off-the-shelf device like hololens or
56:33magic leap or something like that we
56:34even though it cost way more than you
56:36think it's gonna cost from the future
56:37you know this is kind of a trick like
56:40over spend in the early days like spend
56:43more money than you have in order to get
56:45that first test done if it's possible if
56:48you don't have any money then you can't
56:50but if you have a little bit of money
56:52like spend that $400 to get like a
56:57oculus so that you can simulate the
56:59experience in in VR before you actually
57:02go and make the hardware
57:07from your own revenues and profits and
57:10potentially giving so you yeah so I
57:16think I mean I would actually go one
57:17step above that and understand why
57:19hardware companies need financing
57:21usually it's for inventory it's rare
57:24that a company will need to get funding
57:26for it's not rare there's like funding
57:30for inventory and there's funding for
57:32R&D funding for R&D is like nxu like you
57:36need you need that you can either get it
57:38through free payments from potential
57:40customers that's harder but most people
57:42try to raise R&D funding from people at
57:44YC or other investors banks will not
57:46lend you money for R&D your family might
57:50so I would like try to reduce that as
57:53much as possible try to spend as little
57:54money as possible hacking together
57:56prototyping getting to the point where
57:58you know what customer at products
57:59you're going to build then on the
58:01inventory financing side here's the
58:03thing that I've learned inventory sucks
58:07inventory is like one of the hardest
58:09things that our companies have to deal
58:11with because they can't make money
58:13generally until they have a product that
58:15they need to sell it cost money to build
58:18that product and it costs time to build
58:19that product so you have what's called a
58:21cash conversion cycle where you may need
58:23to put inputs dollars into your product
58:26in order to get outputs down the road
58:29again it is impossible it is very
58:33difficult to manage inventory so the
58:36better thing is actually to rise above
58:37it and to get into a situation where you
58:39don't need to buy as much inventory so
58:41there's two ways that you can do this
58:42one is you can get customers to
58:44pre-order that's the easiest way is then
58:47you have a very perfect idea of how many
58:49units you need to build so that you can
58:52and you can use that money from
58:54pre-orders to actually fund the
58:55development you don't need to raise
58:56money from VCS you don't need to raise
58:58money from banks that doesn't work
59:00always but it does work in some crazy
59:01situations like Xiaomi up until two or
59:04three years ago sold their phones on
59:07Tuesday mornings like they had a once
59:10that you couldn't go on the website and
59:12buy xiaomi phones you could only buy
59:14them like Tuesday mornings and so it was
59:15like the thing everyone who would be at
59:17their computers on Tuesday mornings
59:18waiting for like the latest drop Xiaomi
59:21and they had a really short production
59:22cycle not a week but they knew basically
59:25based on these prediction based on these
59:27Tuesday sales how many units they need
59:29to make over the next couple weeks
59:30oneplus also did the same thing so
59:33that's one way of doing it the pre-order
59:35route the other way is to change your
59:38product so that it takes a shorter
59:40amount of time to build this is called
59:42reducing the lead times so the benefit
59:45the secret benefit of reducing your lead
59:47times is that you have to put less of
59:50your hard earned cash into inventory
59:52that will then take months to turn into
59:55actual revenue so interestingly enough
59:58sometimes it's in your best interest to
01:00:02pay more for the product pay more for
01:00:05the bomb but Bill of Materials pay more
01:00:07for the components if the lead times are
01:00:10shorter because you may not actually
01:00:13have a margin problem maybe like you
01:00:16actually have quite a bit of margin free
01:00:18margin but you have this cash flow
01:00:19problem where you can't like fund new
01:00:22batches of products because it takes six
01:00:23months to turn that product into revenue
01:00:25so you could take some of the margin you
01:00:27could do a trade-off where you earn less
01:00:29profit per unit but you're taking less
01:00:31of a risk upfront that's kind of the
01:00:34general learnings that I've had around
01:00:35how didyou Hardware financing great
01:00:39thank you very much Eric thank you